Excuse me, do you have this model?

Dear 业昌

When using the two-phase flow moving mesh interface, then the partition coefficient is used in the boundary conditions. For the concentration for a solute in one phase, e.g. phase 1, becomes a function of the concentration of the same solute in phase 2:
c1 = f(c2)
For the equation of the solute in phase 2, you can instead use the fact that the flux has to be continuous over the phase boundary:
flux(c1)*n1 = – flux(c2)*n2, where n1 and n2 are the normal outwards vector at the boundary for the respective phase. The principle is demonstrated for a membrane in this model, see specifically slide 5 in the PowerPoint file:
https://www.comsol.com/model/separation-through-dialysis-258

For a dispersed flow model, this becomes more complicated to formulate but on the other hand it is also easier to solve. You then need to have an expression for the mass transfer coefficient at the phase boundary, usually only in the phase where transport is rate determining. This may also yield an additional concentration variable at the phase boundary for each phase and each solute. Then you use the same principle as above but for the homogenized reaction term instead, i.e. for one solute equation you use the equilibrium condition at the boundary and for the other you use the mass transport coefficient to get an equal reaction but with opposite sign in each point in space for the two phases. We will create a few example models with this feature exemplified during the year.

Best regards,
Ed